Reality Check: Challenges with Reporting Data from Various Sources
Dr. Nisha Thampi, MD, MSc
Medical Director, Infection Prevention and Control Program
Children’s Hospital of Eastern Ontario, Ottawa
May 26, 2019
Disclosure Statement
• I have no affiliation (financial or otherwise) with a pharmaceutical, medical device or communications organization.
Measurement is the first step that leads to control and eventually to improvement. - H. James Harrington
Measure understand control improve
But…
Pressure on a program to measure inaccurately so the problem remains poorly understood
Individuals may perceive and therefore measure healthcare-associated infections differently
And…
Critical to follow standard definitions and ensure consistency
• Ascertainment of infection
• Attribution to healthcare setting
• Reporting internally and externally
HAI data under scrutiny by public payer as quality metrics(Reimbursement in the US)
Objectives
• Review process and outcome measures in surveillance of healthcare-associated infections
• Review impact of reportable HAI surveillance indicators
• Highlight pitfalls in accurately reporting data from various sources
Objectives
• Review process and outcome measures in surveillance of healthcare-associated infections
• Review impact of reportable HAI surveillance indicators
• Highlight pitfalls in accurately reporting data from various sources
Value of surveillance
• Systematic, ongoing collection and analysis of data during a defined period of time using standardized definitions• May also involve laboratory confirmation and chart review
• Timely dissemination of information to those who require it in order to take action = improve patient safety
• Why?• Provides measure of burden of illness• Establishes benchmark rates for internal and external comparison
• May motivate hospitals to adhere more closely to best practices in infection prevention
• Identifies potential risk factors• Allows for assessment of specific interventions
PIDAC-IPC 2014: Best Practices for Surveillance of Health Care-associated Infections in Patient and Resident Populations
HAI surveillance indicators: outcomes
• Communicable diseases and events
• Body site-specific• Central line-associated bloodstream infections• Surgical site infections• Infection-related ventilator-associated events• Catheter-associated urinary tract infections
• Organism-specific healthcare-associated infections• Methicillin-resistant Staphylococcus aureus• Vancomycin-resistant Enterococci• Carbapenemase-producing organisms• C.difficile
HAI surveillance indicators: processes
• Central line-associated bloodstream infections• Hand hygiene prior to accessing central line• Bundle adherence (5-7 elements)
• Surgical site infections• Adherence to pre-operative MRSA screening protocol• Appropriateness of surgical antimicrobial prophylaxis
• Infection-related ventilator-associated events• Adherence to practices that reduce infection risk associated with ventilator
use
• Catheter-associated urinary tract infections• Adherence to practices to limit urinary catheter use
Hand hygiene!
Objectives
• Review process and outcome measures in infection prevention and control
• Review impact of reportable HAI surveillance indicators
• Highlight pitfalls in accurately reporting data from various sources
“hospital-wide surveillance of hospital-acquired infections provides appropriate targets for interventions tailored to the specific needs of the hospital”
Surveillance associated with 30% decrease in rate of pneumonias and SSI, 20% decrease in BSI
Journal of the Pediatric Infectious Diseases Society 2018;7(1):18–24
Prospective surveillance for HA-VRI found higher incidence rates compared to hospital-associated bloodstream infections
Organized, intensive surveillance and control activities associated with 32% decrease in HAI
Following introduction of public reporting, C. difficile infections declined by 26% across Ontario = >1,900 cases averted per year
Collaboration in surveillance + bundle implementation= 21% reduction in SSI rate of pediatric cardiothoracic, neurosurgical ventricular shunt, and spinal fusion surgeries
Canadian Nosocomial Infection Surveillance Program
• Established in 1994
• Epidemiology and molecular characteristics of healthcare-associated infections in Canadian hospitals
• Partnerships: PHAC, NML, AMMI
• 73 sentinel hospitals in 10 provinces• 9 freestanding pediatric hospitals
• ~ 78% of Canadian population lives within 100km of CNISP site
15
CNISP
• Provides Canadian hospitals with “benchmark” data • Standardized HAI surveillance case definitions - reviewed annually
• National, regional and site-specific HAI rates, strain types and antimicrobial resistance and utilization data
• Provides evidence-based data:• Antimicrobial surveillance program (CARSS)
• Canadian infection prevention and control guideline preparation (National Advisory Committee for Infection Prevention and Control)
• Raises public awareness of important infection control issues relating to AROs and HAIs
https://www.canada.ca/en/public-health/services/infectious-diseases/nosocomial-occupational-infections.htmlhttps://ipac-canada.org/cnisp-publications.php
Objectives
• Review process and outcome measures in infection prevention and control
• Review impact of reportable HAI surveillance indicators
• Highlight pitfalls in accurately reporting data from various sources• Case study
Variability in reporting
• What: clinical diagnosis vs standard definition
• Who: infection control professional, coders, clinicians in admin roles
• When: continuously or periodically
• Where: internal and external reporting authorities
• How are data reported: numerators and denominators
What is reported
• Sources of surveillance data:• Patient chart/records (e.g. pharmacy, medical imaging)
• Laboratory reports
• Safety reports
• Clinical rounds
• Communication with caregivers
• Decision Support (coders)
• Surveillance-specific forms (e.g. NICU patients with central lines by weight category)
• Surgical Information System
• Critical Care Information System
What is reported
• Clinical diagnosis vs surveillance definitions
• Critically important to have uniform application of standardized and validated case definitions• For both outcome and process measures
• Unique challenges to pediatric surveillance:• Surgical site infection
• ICD-10 codes do not have specific codes for pediatric procedures e.g. Tetralogy of Fallot
• Contentious diagnosis (“purulent drainage” vs “incision clean and dry” by MDs)
• C.difficile infection
Case: Clostridium difficile reporting
• Colonization is common in children under 2 yo• Higher rates of colonization, lower rates of complications compared to adults
• Benchmarked against adult teaching and community hospitals in provincial reporting
• IDSA “strongly discourages” public reporting of cases in children <2 yo• CNISP and most provincial ministries of health report rates among 1-18 yo
• Alberta moving to 2-18 yo
= What gets measured as a quality metric?
• Case definition for CNISP reporting:• 3 x loose, watery stools or fever, abdominal pain and/or ileus• “without reasonable evidence of another cause of diarrhea”
• >70% cases found to have alternate pathogen present
= Requires individual chart review + lab confirmation
Clinical Practice Guidelines for Clostridium difficile infection: IDSA 2018AMMI Canada treatment practice guidelines for Clostridium difficile: 2018
When are HAI data reported?
• Continuously or concurrently
• Periodically (retrospective)
• Challenges• Human resources for case finding and review
• Different case definitions among reporting bodies
• May not be able to determine “truth” when retrospectively reviewing data due to numerous sources
Who does the reporting?
• Infection control professional
• Coder from Decision Support
• Clinical team: notifies IPAC to review if clinical suspicion of HAI
• Clinical nurse: temporary administrative role to support QI initiative
Where do HAI data go?
• Internal committee: IPAC, Quality and Safety; Executive, Board
• Provincial mandatory reporting
• Quality improvement / surveillance programs• Solutions for Patient Safety (US and Canadian network of pediatric hospitals)
• National Surgical Quality Improvement Program
• Canadian Nosocomial Infection Surveillance Program
How are HAI reported?
• Numerators: • Based on case definition; can differ between surveillance programs
• e.g.: CSF shunt infection for CNISP = up to 1 year post-op • SPS = up to 90 days post-op
• We can create workflow process, but which rate to report internally?
• Denominators:• Not consistent between surveillance programs
Denominator NSQIP SPS CNISP
Cardiovascular surgery surveillance No Yes Yes
General surgery surveillance Yes No No
Pericardial window procedures No Yes No
Case Study:
• Hospital IPAC program prospectively identifies HAIs• Submits data to CNISP, Ministry of Health, Critical Care Services Ontario
• 2016: Hospital joined 2 surveillance networks:• Solutions for Patient Safety to reduce risk of CLABSI, SSI
• CV surgery requiring bypass, spinal fusion, neurosurgical shunt infections
• IPAC program central to surveillance strategy and development of process audits
• National Surgical Quality Improvement Program-Pediatrics network• Data abstractor hired by NSQIP team to prospectively collect preoperative,
intraoperative and postoperative data, with outcomes at 30 days after index surgery• Surveillance every 8 days x 35 cases, no dental or CV surgery
• Data submitted directly to NSQIP and anonymized in network publications
• Hospital benchmarked against similar institutions, info sent to hospital leadership
Case Study:
• 2017: CIHI approached hospital in advance of launching “In-hospital infections indicators”• 2 publicly reportable indicators: MRSA, C.difficile
• Meant to complement Patient Safety indicators (e.g. in-hospital sepsis) with site-specific data
• Infection data to be collected by coders
• Coders review clinical chart and code for HAI if documented in physician’s note
Case study: Current state
• CNISP• ICP collects data
• Most of CNISP data reported to Solutions for Patient Safety, MOH
• NSQIP: • Reported directly to NSQIP and internally by clinical administrator
• ICP may receive data to validate, with short turnaround time
• Solutions for Patient Safety• ICP collects data, sends to clinical administrator to report internally and to SPS
• CIHI• No published HAI indicator
Challenges = human resource-intensive; various denominators for same measureWhich one reflects the truth? Which one to follow in quality improvement?
• Case studies published in AJIC with link to online survey• Multiple-choice questions based on standard surveillance criteria and protocols
• Assessed accuracy and consistency in applying standardized surveillance definitions
• 62.5% responses correct • ICPs responded correctly (62%) significantly more often than physicians (55%)
• Highlights need for continuing education, competency development, auditing
How to overcome challenges in data quality?
• One source of information internally and externally• All data collected by IPAC
• Workload? Opportunity costs?
• All data reviewed by IPAC• Forum (if no IPAC Committee)? Frequency? Responsibility if inaccurate data collection?
• HAI case finding and reporting• How to assess competency in applying case definitions within the
organization?
• How to measure accuracy of reporting
Acknowledgements
• Pat Bedard, RN, CIC
• Allyson Shephard, RN, CIC